ICON-eCLM-ParFlow: Treatment of lakes #116

Migrated from https://gitlab.jsc.fz-juelich.de/HPSCTerrSys/tsmp-internal-development-tracking/-/issues/28

Originally posted by @s-poll last 11-May-2023

I want to open this issue for discussion on the default treatment of lakes in the (new) fully coupled TSMP system. The default treatment of the COSMO-CLM3.5-ParFlow coupling was that lakes are inactive in CLM3.5, and using the integrated treatment of lakes for COSMO. To my knowledge, the lakes were inactive in ParFlow, but were switch to active grid points to ensure river outflow of some lakes for new domains e.g. in DETECT (please correct in case). Technically, eCLM can be coupled to ICON with all landunits including the lake landunit. I do not see a chance to include a dynamical lake fraction based on pounding water of ParFlow.

CLM5.0 uses the Lake, Ice, Snow, and Sediment Simulator (LISSS) schemata, while ICON uses the FLake schemata. Based on a first evaluation the FLake schemata is based on self‐similarity, while the LISSS schemata is wind‐driven eddy diffusion. In this publication you can find a comparison of the scheme for a case study.

Shell we use active Lake points in eCLM? If yes, should we couple these points to ICON/ParFlow?